For the Taisaw filter, the passband bandwidth is quoted as 5 MHz, but looking at the response curve (S21), it looks like it could pass about 75 to 90 MHz bandwidth without much attenuation. Likewise for the Vectron filter, the passband bandwidth is quoted as 2.4 MHz, but S21 suggests that it can actually pass around 40 MHz bandwidth without much attenuation.
So how is passband bandwidth defined? Perhaps based on S11?
Bandwidth is usually -6db, according to S21 that is about roughly 90Mhz.
But I suspect only the very centre 5Mhz is of real interest, where very good figures for VSWR and passband ripple are quoted.
What I think all this means is that this is an initial filter to block strong out of band signals that may cause problems in a following mixer.
Further down the chain at much lower IF frequency a much more selective filter will be employed.
So although it is pretty broad at 90 Mhz, the specifications over the required 5Mhz used are of particular interest for some application.
For the Taisaw filter, set reference without filter, put in filter, this then gets you your insertion loss (1.7 dB) measure bandwidth at 3dB points = 5 MHZ. The bandwidth at -6dB is 8.5 MHZ. At 80 MHZ band width the loss is 20 dB, or 10% of your signal gets through it. So it depends on how you define " without much attenuation.".
Frank
As far as I see, many of saw manufacturers do not provide clear data. You may need to look for graphs and deduce the bandwidth info from the graphs on your own. Furthermore, you may have to ask for s-parameters from the manufacturer to have information you need. Taisaw accepts s-parameter requests.
Are you looking at the correct datasheet? I dont see how u got -6db at 8.5 MHz....
Regards
Aaron
- - - Updated - - -
Thanks for all the replies. Normally im used to seeing -3dB, -6dB etc bandwidths, but it looks like there may be another standard too...just wondering if it is a common standard, and what it is...